Acceso usuarios
[1]
This article deals with the realization of referential subjects in the L2 Spanish of German (adult) native speakers. The acquisition of a null subject grammar by speakers of a non-null subject language has drawn considerable attention in generative approaches to L2 acquisition. This article revisits the issue and compares the predictions made by the Interface Hypothesis (Sorace 2005, 2011, Sorace and Filiaci 2006, Tsimpli and Sorace 2006) to an alternative, the Feature Reassembly Hypothesis (Lardiere 2008, Slabakova 2013 Cho and Slabakova 2014). Relying on corpus data, the study presents a novel empirical approach and applies an innovate statistical analysis procedure from learner corpus research. The results of the study corroborate previous empirical findings, namely that pronouns, yet not null subjects, are problematic, but also brings in new insights, in particular that issues with pronouns are consistent and go beyond the contexts predicted by the Interface Hypothesis. The contrasts between L1 and L2 subject realization found in the data therefore can only in part be explained to result from interface issues. The Feature Reassembly Hypothesis offers a suitable additional explanation relating the issues to the properties of the L1 and L2 learnability.
